毕业论文：《基于BP神经网络对赣江水质的评测》.doc

东华理工大学毕业设计(论文) 摘要 PAGE I 毕业设计（论文） 题 目:BP神经网络对赣江水质的评价 英文题目:BP neural network to Ganjiang 学生姓名： 学 号： 专 业: 软件工程 学 院： 软件学院 指导教师： 职称： 助教 二零一一年五月 摘 要 人工神经网络以人脑结构为参考模型，试图通过简单计算但也的高速互联，来实现类似于人类在语言和图像处理等方面的行为。它是由简单信息处理单元互联组成的网络，能够接受并处理信息。网络的信息处理是由处理单元之间的相互作用来实现的。 基于BP神经网络的水质评价模型的训练样本即为水质分级标准。训练完成后，网络将保存对分级标准学习的知识和有关信息，即各层间的连接权与各个神经元的阈值得以保存，然后从输入层输入待评资料，得到有关评价结论的信息，从而根据一定规则作出有关的评价结论的判断。 本文运用BP神经网络理论和方法，使用MATLAB 工具箱函数编程，建立了赣江水质综合评价的模型，对赣江近9年的监测水质进行了评价。评价结果显示，赣江水质总体上集中在Ⅱ类水质上，属于较清洁水；同时，评价结果也表明，BP 神经网络可以较好地实现水质综合评价，且具有较高的实用性和客观性，完全可以应用于实际的水质综合评价工作。 关键词：BP神经网络；水质评价；赣江；MATLAB 东华理工大学毕业设计(论文) ABSTRACT II ABSTRACT The manual nerve network takes person brain structure as and makes reference to pattern and tries and passes a simple computing but also of high speed with each other allied, carry out the behavior that analogy handles and waits aspect in the language and the ICON at the mankind.It is united a composite network with each other by the simple information handing unit, accept and handle an information.The information handing of network is carried out by the interaction of of transaction unit. It fored a fluid matter to graduate standard according to the training sample of the fluid matter appraisal pattern of BP nerve network.After training to complete, network conservancy vs the knowledge that graduates standard study and relevant information, namely the Yu of each nerve kyat of connecting right with of each worth with keep, then the input need to be reviewed data from the input layer and receive concerning the appraisal conclusion of information, thus make the criterion of a concerned appraisal conclusion according to the certain rule. This text usage BP nerve network theory and method, the tool box function of use MATLAB weaves a distance and created Gan river fluid matter synthe